I have a question about mixing data values and flag values. I know the correct
method is to separate the data information from the flag information, but my
colleague would prefer to keep the two in the same variable to simplify
processing and reduce file size. The data files are already quite large. I
looked through the discussion archive but could not find any discussion about
this.
For example:
I have a data vector that occasionally has missing value indicators. They want
to use multiple missing value indicators to indicate why the data is missing. A
value of -9999 indicates no data at all, a value of -2 indicates possible clear
sky, -1 indicates it is clear sky, and anything greater than or equal to 0
represents a data value. We will use valid_range to indicate that any value
below 0 is not valid data and will be excluded. (I can’t use valid_min for
historical reasons. We are sticking with missing_value for historical reasons.)
cloud_layer_base_height(time, layer):float
long_name = "Cloud base height of hydrometeor layers” ;
units = "m” ;
missing_value = -9999.f ;
valid_range = 0.f, 100000.f ;
flag_values = -2.f , -1.f ;
flag_meanings = "possible_clear_sky clear_sky” ;
comment = "-2. Possible clear sky (No MPL observations available,
Ceilometer obscured, but no cloud detected), -1. Clear sky, >= 0. Valid cloud
base height” ;
Would this be acceptable?
Ken
Kenneth E. Kehoe
Research Associate
ARM Climate Research Facility Data Quality Office
Office: 303-497-4754
Cell: 405-826-0299
_______________________________________________
CF-metadata mailing list
[email protected]
http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
